The market for grants
Manifund helps great charities get the funding they need. Discover amazing projects, buy impact certs, and weigh in on what gets funded.
Chris Wendler
3 days ago
I want to put my money where my mouth is and support my student's conference trip.
Joel Becker
4 days ago
Donated another $20k for $40k total.
I think it's crazy that we live in a world where AIs are well on their way to saturating ~all challenging technical benchmarks and yet they can't pull off basic experiments. It seems to me that exploring this harder would be very informative for learning about the capability profile of AIs. Right now it's unclear what we should learn from AI village because (1) I don't know how much I should trust quality of the set-up/elicitation, and (2) the effective sample size is tiny.
One blocker for this happening is limited staff time. Another blocker -- totally speculative on my part, but my guess is shared between both other grantmakers and Sage/AI digest staff themselves -- is viewing AI digest as primarily a comms play and not a science play. I think AI village should primarily be seen as a science play (that currently isn't doing a great job at science but has a tonne of potential).
I know @Austin is a bit concerned that this grant opportunity is not Manifund's comparative advantage/can easily be considered by other funders. I disagree -- it seems to me that traditional funders (and perhaps the staff behind AI village themselves) are being far too unambitious/unexcited about this project (on the basis of it being a comms play with limited existing comms traction), and I'm thrilled to put some of my Manifund pot towards nudging this situation in a better direction.
Thane Ruthenis
5 days ago
What are the largest sources of uncertainty for this project, or reasons it might fail to have impact?
The main failure modes are:
Failure of the core idea: that we could recover an interpretable model of the world by some easy-to-specify training/search process.
The underlying assumptions seem fairly solid to me:
We have many sources of evidence, empirical and theoretical, that the universe has an "observer-independent" well-abstracting structure.
The property of being "easy-to-interpret" is not inherently "human-laden". The process doesn't have to locate the human ontology and translate the world-model into it, it just has to structure the world-model in a fairly natural way that'd make it easy for us to untangle/study.
Approaching the problem via compression likewise seems well-founded, including empirical work pointing in that direction (see e. g. this and that).
That said, the project tackles some nontrivial theoretical problems, and it's possible that none of this works quite the right way.
Excessively long R&D time. Even if the core idea is correct, the time to flesh out the theory and cross the theory-practice gap may exceed the time it'd take others to achieve AGI by other means.
I aim to mitigate this by aggressively prioritizing R&D avenues that promise to connect to reality quickly, and by using existing well-developed methods to implement whichever parts of the project could be implemented that way, instead of reinventing the wheel in the name of theoretical elegance. (My current expectation is that all the tools are already there, theory is only needed to figure out how to arrange them the right way.)
In addition, as I'd mentioned, worlds in which AI progress is rapid are worlds in which we can probably use that progress' byproducts (e. g., superhuman-mathematician LLMs) to dramatically speed up research on this project as well. On the other hand, worlds in which AI progress isn't as rapid as the industry hopes are worlds in which we can afford the longer development time.
It's still possible it'd all take longer than expected.
What other sources of funding have you sought?
I've also applied to the LTFF. (Though I expect it'd take a couple months to hear back, and my application prioritized smaller-but-more-certain funding, so I don't expect to be exhaustively funded by it.)
Ryan Kidd
5 days ago
Hey Thane! What other sources of funding have you sought? What are the largest sources of uncertainty for this project, or reasons it might fail to have impact?
Francisco Carvalho
7 days ago
Our Community Archive browser extension is live! It unlocks a stream of tweets in real time.
Introducing epistemic.garden, an R&D lab for community-driven collective intelligence!
Just as cells don’t realize they’re part of larger beings, we don’t always see the egregores and memeplexes shaping us. Our strategy doc lays out how we’re building tools to make these dynamics visible.
In our last update, I asked for support to start an R&D lab that builds tools to map how ideas spread.
We were graciously supported by Vitalik Buterin, the Survival and Flourishing Fund, and Peter Wang! Thank you patrons 🙏
I wrapped up my job at the EU AI Office that I started in the mean time and finished some travel.
Now we’re back on track!
We’ll be working with strategic partners to build prototypes like a memetic weather report (‘what topics were up last week?’) and an embedding atlas that maps the terrain - communities of users.
NYC October: I’ll be in New York Oct 8 – Nov 4 and plan to host a gathering on collective sensemaking, memetics, and AI-for-epistemics. If you’re around, I’d love to see you there.
Warmly,
xiq
Connor Axiotes
8 days ago
Thank you, Michael. These are very fair questions! Mike has answered them here below:
The total required to complete Making God is $1,400,000. We have received $350,000 so far through a mix of donations and investments. We have a further $400,000 promised in private investment. That leaves $650,000 to raise to finish production, complete post, and run an international festival and sales campaign. The $1.4M total includes the festival and sales budget, including the previously referenced $102,000 festival route line.
On the 2023 UK budget survey: it covers everything from micro-budget indies to broadcaster features. Our comparison set is feature documentaries intended for international distribution, with multi-country shoots, legal clearances, festival deliverables, and streamer-ready masters. Those commonly land in the low seven figures. Our $1.4M target is disciplined for that category. The earlier $730,000 figure was a minimum route to a festival-only cut. We are budgeting for a distribution-grade film because that is the appropriate path to broad impact.
We have spent about $300,000 to date. That has produced more than 50 hours of filmed material across multiple countries, including interviews with Geoffrey Hinton, Will MacAskill, and Gary Marcus.
Since March 2025, we've have taken $70,000 combined in salaries and have deferred a further $40,000 until investment closes. Our day rates are below union benchmarks and lower than other key crew. The remainder of spend is production: crew fees, rentals, insurance, legal, travel and accommodation, and production administration. Any equipment purchased is owned by the company and recorded on the asset register. No personal assets were bought with production funds.
Why this split: the value sits on screen and in a clean rights position for sale. Most spend therefore goes to filming, travel, crews, and lawful clearances. Modest pay for the core team keeps the project moving across months of prep, shoots, and edit, and we have kept that lean with deferrals to protect runway.
Going forward, donated funds are used for production and post costs that benefit the film directly. Producer and director compensation in post is planned to be covered by private investment.
Tail End Films Limited will be the production company. A dedicated SPV will be created soon holding the Making God IP and contract with investors, which gives buyers a clean chain of title.
We will seek to sell the film for more than it cost to make, while being candid that documentary outcomes are uncertain. Investors participate under a standard recoupment and revenue-share waterfall, administered to buyer requirements and, if appropriate, through a collection account manager. To avoid any perceived mismatch between donors and investors, we have invited large donors to convert to the investment round on the same terms as other investors, with a sensible minimum ticket of $10,000. Donors who remain donors keep their credit and we report impact transparently.
The general public has not been widely exposed to figures like Geoffrey Hinton and Yoshua Bengio, and when they are, the material is rarely presented in a way that is both engaging and clear. Our job is to translate credible concerns into a compelling, accessible story for a broad platform audience. We will announce additional contributors once clearances are in place, and we only include claims that are corroborated and legally cleared.
On the early behind the scenes clips you saw: those were internal sizzles. In the film we appear only when it adds stakes or clarity. Otherwise the audience stays with the material.
We are aiming for a streamer-ready feature documentary. That means resourcing the edit properly and budgeting for deliverables, rights, and quality control that buyers actually require.
Post production studio. Ensures the cut is properly resourced and delivered on schedule with QC, subtitles, and delivery elements buyers require.
Archival licensing. For worldwide, all media distribution, long-term rights across territories cost more than festival-only use, so we budget up front rather than weaken the film later.
Stock B-roll reserve. Used only where filming is impractical and a licensed shot completes a sequence. We prefer to film ourselves and will commission remote shooting where that is the smarter option.
Motion graphics and design. Builds a clear visual system so general audiences can follow complex ideas. Mike leads creative and does hands-on work to keep costs efficient.
Composer and musicians. Secures a quality score with stems and cue sheets on terms a buyer accepts. Where a composer prefers some back end, the cash fee can reduce.
Contingency. A prudent buffer for multi-vendor work and international clearances. If it is not needed, it is not spent.
These are ceilings, not targets. We will lock fixed bids where possible and redirect savings to the cut or reduce the top line.
SB-1047 short. Mike handled motion graphics and editing. About ten working days of art direction was typical for the scope and budget.
Campaign totals. The 60M figure is cumulative across platforms while Mike was Creative Director at Control AI. The 32M figure refers to the cross-platform performance of Your Identity Isn’t Yours. That campaign had significant paid media by design, roughly half of its impressions, and it produced outcomes including more than 3,000 letters to UK MPs and coverage on ITV News. We can share per-platform analytics that separate paid and organic performance on request.
Collaboration with Max Tegmark. Mike collaborated on a deepfake demo at Davos in January 2024. Coverage: https://www.tortoisemedia.com/2024/01/18/davos-elite-worry-about-deepfakes-and-disinformation
Apollo Research. A three-week engagement, one week of pre-production and two weeks to deliver a motion-graphics explainer of their scheming research. The video is on Apollo Research’s YouTube channel.
Commercials. Starbucks was a B2B spot with Pyramid Productions for franchisees, so it is not publicly hosted. Mike can provide the production contract. Pale Waves was a lyric video for She’s My Religion that ran on socials; a short snippet is on Mike’s Instagram. Mandarin Oriental was an early social commission, example: https://www.instagram.com/reel/CHNy_ocnTPT/
Long form directing. This unfortunately was a hallucination by an AI model we used to edit the Manifund project copy. While Mike has been directing Making God since March 2024, he has not directed any other documentaries before this one. We will endeavour to not make that mistake again!
We keep a detailed cost ledger with receipts and reconcile monthly. We work with legal counsel on releases, music, archival, and E and O insurance. Donated funds are ring-fenced to production and post costs. Investor funds flow through the SPV that holds the IP, with standard recoupment and reporting. We clear music and archival on worldwide, all media, in perpetuity terms, bind E and O prior to delivery, and maintain signed releases and a full chain of title.
If anything here needs further detail, Mike is happy to share a more granular cost breakdown and, when legal and licensing allow, private screeners of selected scenes. @michaeltrazzi
Benjamin Sturgeon
13 days ago
I have a conflict of interest as someone who is a strategic director on the project, but from the level of excitement about the project from people I've spoken to who are interested in applying and my knowledge of the team I believe that the project has a lot of value to offer.
Chris Wendler
3 days ago
I want to put my money where my mouth is and support my student's conference trip.
Joel Becker
4 days ago
Donated another $20k for $40k total.
I think it's crazy that we live in a world where AIs are well on their way to saturating ~all challenging technical benchmarks and yet they can't pull off basic experiments. It seems to me that exploring this harder would be very informative for learning about the capability profile of AIs. Right now it's unclear what we should learn from AI village because (1) I don't know how much I should trust quality of the set-up/elicitation, and (2) the effective sample size is tiny.
One blocker for this happening is limited staff time. Another blocker -- totally speculative on my part, but my guess is shared between both other grantmakers and Sage/AI digest staff themselves -- is viewing AI digest as primarily a comms play and not a science play. I think AI village should primarily be seen as a science play (that currently isn't doing a great job at science but has a tonne of potential).
I know @Austin is a bit concerned that this grant opportunity is not Manifund's comparative advantage/can easily be considered by other funders. I disagree -- it seems to me that traditional funders (and perhaps the staff behind AI village themselves) are being far too unambitious/unexcited about this project (on the basis of it being a comms play with limited existing comms traction), and I'm thrilled to put some of my Manifund pot towards nudging this situation in a better direction.
Thane Ruthenis
5 days ago
What are the largest sources of uncertainty for this project, or reasons it might fail to have impact?
The main failure modes are:
Failure of the core idea: that we could recover an interpretable model of the world by some easy-to-specify training/search process.
The underlying assumptions seem fairly solid to me:
We have many sources of evidence, empirical and theoretical, that the universe has an "observer-independent" well-abstracting structure.
The property of being "easy-to-interpret" is not inherently "human-laden". The process doesn't have to locate the human ontology and translate the world-model into it, it just has to structure the world-model in a fairly natural way that'd make it easy for us to untangle/study.
Approaching the problem via compression likewise seems well-founded, including empirical work pointing in that direction (see e. g. this and that).
That said, the project tackles some nontrivial theoretical problems, and it's possible that none of this works quite the right way.
Excessively long R&D time. Even if the core idea is correct, the time to flesh out the theory and cross the theory-practice gap may exceed the time it'd take others to achieve AGI by other means.
I aim to mitigate this by aggressively prioritizing R&D avenues that promise to connect to reality quickly, and by using existing well-developed methods to implement whichever parts of the project could be implemented that way, instead of reinventing the wheel in the name of theoretical elegance. (My current expectation is that all the tools are already there, theory is only needed to figure out how to arrange them the right way.)
In addition, as I'd mentioned, worlds in which AI progress is rapid are worlds in which we can probably use that progress' byproducts (e. g., superhuman-mathematician LLMs) to dramatically speed up research on this project as well. On the other hand, worlds in which AI progress isn't as rapid as the industry hopes are worlds in which we can afford the longer development time.
It's still possible it'd all take longer than expected.
What other sources of funding have you sought?
I've also applied to the LTFF. (Though I expect it'd take a couple months to hear back, and my application prioritized smaller-but-more-certain funding, so I don't expect to be exhaustively funded by it.)
Ryan Kidd
5 days ago
Hey Thane! What other sources of funding have you sought? What are the largest sources of uncertainty for this project, or reasons it might fail to have impact?
Francisco Carvalho
7 days ago
Our Community Archive browser extension is live! It unlocks a stream of tweets in real time.
Introducing epistemic.garden, an R&D lab for community-driven collective intelligence!
Just as cells don’t realize they’re part of larger beings, we don’t always see the egregores and memeplexes shaping us. Our strategy doc lays out how we’re building tools to make these dynamics visible.
In our last update, I asked for support to start an R&D lab that builds tools to map how ideas spread.
We were graciously supported by Vitalik Buterin, the Survival and Flourishing Fund, and Peter Wang! Thank you patrons 🙏
I wrapped up my job at the EU AI Office that I started in the mean time and finished some travel.
Now we’re back on track!
We’ll be working with strategic partners to build prototypes like a memetic weather report (‘what topics were up last week?’) and an embedding atlas that maps the terrain - communities of users.
NYC October: I’ll be in New York Oct 8 – Nov 4 and plan to host a gathering on collective sensemaking, memetics, and AI-for-epistemics. If you’re around, I’d love to see you there.
Warmly,
xiq
Connor Axiotes
8 days ago
Thank you, Michael. These are very fair questions! Mike has answered them here below:
The total required to complete Making God is $1,400,000. We have received $350,000 so far through a mix of donations and investments. We have a further $400,000 promised in private investment. That leaves $650,000 to raise to finish production, complete post, and run an international festival and sales campaign. The $1.4M total includes the festival and sales budget, including the previously referenced $102,000 festival route line.
On the 2023 UK budget survey: it covers everything from micro-budget indies to broadcaster features. Our comparison set is feature documentaries intended for international distribution, with multi-country shoots, legal clearances, festival deliverables, and streamer-ready masters. Those commonly land in the low seven figures. Our $1.4M target is disciplined for that category. The earlier $730,000 figure was a minimum route to a festival-only cut. We are budgeting for a distribution-grade film because that is the appropriate path to broad impact.
We have spent about $300,000 to date. That has produced more than 50 hours of filmed material across multiple countries, including interviews with Geoffrey Hinton, Will MacAskill, and Gary Marcus.
Since March 2025, we've have taken $70,000 combined in salaries and have deferred a further $40,000 until investment closes. Our day rates are below union benchmarks and lower than other key crew. The remainder of spend is production: crew fees, rentals, insurance, legal, travel and accommodation, and production administration. Any equipment purchased is owned by the company and recorded on the asset register. No personal assets were bought with production funds.
Why this split: the value sits on screen and in a clean rights position for sale. Most spend therefore goes to filming, travel, crews, and lawful clearances. Modest pay for the core team keeps the project moving across months of prep, shoots, and edit, and we have kept that lean with deferrals to protect runway.
Going forward, donated funds are used for production and post costs that benefit the film directly. Producer and director compensation in post is planned to be covered by private investment.
Tail End Films Limited will be the production company. A dedicated SPV will be created soon holding the Making God IP and contract with investors, which gives buyers a clean chain of title.
We will seek to sell the film for more than it cost to make, while being candid that documentary outcomes are uncertain. Investors participate under a standard recoupment and revenue-share waterfall, administered to buyer requirements and, if appropriate, through a collection account manager. To avoid any perceived mismatch between donors and investors, we have invited large donors to convert to the investment round on the same terms as other investors, with a sensible minimum ticket of $10,000. Donors who remain donors keep their credit and we report impact transparently.
The general public has not been widely exposed to figures like Geoffrey Hinton and Yoshua Bengio, and when they are, the material is rarely presented in a way that is both engaging and clear. Our job is to translate credible concerns into a compelling, accessible story for a broad platform audience. We will announce additional contributors once clearances are in place, and we only include claims that are corroborated and legally cleared.
On the early behind the scenes clips you saw: those were internal sizzles. In the film we appear only when it adds stakes or clarity. Otherwise the audience stays with the material.
We are aiming for a streamer-ready feature documentary. That means resourcing the edit properly and budgeting for deliverables, rights, and quality control that buyers actually require.
Post production studio. Ensures the cut is properly resourced and delivered on schedule with QC, subtitles, and delivery elements buyers require.
Archival licensing. For worldwide, all media distribution, long-term rights across territories cost more than festival-only use, so we budget up front rather than weaken the film later.
Stock B-roll reserve. Used only where filming is impractical and a licensed shot completes a sequence. We prefer to film ourselves and will commission remote shooting where that is the smarter option.
Motion graphics and design. Builds a clear visual system so general audiences can follow complex ideas. Mike leads creative and does hands-on work to keep costs efficient.
Composer and musicians. Secures a quality score with stems and cue sheets on terms a buyer accepts. Where a composer prefers some back end, the cash fee can reduce.
Contingency. A prudent buffer for multi-vendor work and international clearances. If it is not needed, it is not spent.
These are ceilings, not targets. We will lock fixed bids where possible and redirect savings to the cut or reduce the top line.
SB-1047 short. Mike handled motion graphics and editing. About ten working days of art direction was typical for the scope and budget.
Campaign totals. The 60M figure is cumulative across platforms while Mike was Creative Director at Control AI. The 32M figure refers to the cross-platform performance of Your Identity Isn’t Yours. That campaign had significant paid media by design, roughly half of its impressions, and it produced outcomes including more than 3,000 letters to UK MPs and coverage on ITV News. We can share per-platform analytics that separate paid and organic performance on request.
Collaboration with Max Tegmark. Mike collaborated on a deepfake demo at Davos in January 2024. Coverage: https://www.tortoisemedia.com/2024/01/18/davos-elite-worry-about-deepfakes-and-disinformation
Apollo Research. A three-week engagement, one week of pre-production and two weeks to deliver a motion-graphics explainer of their scheming research. The video is on Apollo Research’s YouTube channel.
Commercials. Starbucks was a B2B spot with Pyramid Productions for franchisees, so it is not publicly hosted. Mike can provide the production contract. Pale Waves was a lyric video for She’s My Religion that ran on socials; a short snippet is on Mike’s Instagram. Mandarin Oriental was an early social commission, example: https://www.instagram.com/reel/CHNy_ocnTPT/
Long form directing. This unfortunately was a hallucination by an AI model we used to edit the Manifund project copy. While Mike has been directing Making God since March 2024, he has not directed any other documentaries before this one. We will endeavour to not make that mistake again!
We keep a detailed cost ledger with receipts and reconcile monthly. We work with legal counsel on releases, music, archival, and E and O insurance. Donated funds are ring-fenced to production and post costs. Investor funds flow through the SPV that holds the IP, with standard recoupment and reporting. We clear music and archival on worldwide, all media, in perpetuity terms, bind E and O prior to delivery, and maintain signed releases and a full chain of title.
If anything here needs further detail, Mike is happy to share a more granular cost breakdown and, when legal and licensing allow, private screeners of selected scenes. @michaeltrazzi
Benjamin Sturgeon
13 days ago
I have a conflict of interest as someone who is a strategic director on the project, but from the level of excitement about the project from people I've spoken to who are interested in applying and my knowledge of the team I believe that the project has a lot of value to offer.